Speaking at Palo Alto Networks 2015 Ignite Conference last week, Columbia Sportswear global IS communications manager John Spiegel revealed the companys' aims to create a ‘software defined data centre’ using technologies such as VMware’s NSX.
The increasing automation of infrastructure and abstraction of control to the software layer has been a hot topic for some time now. Here are some of his tips for how network engineers can adapt to the rapidly changing future of the data centre...
1. 1. Prepare for disruption
“Many of you are out on the frontline, configuring hardware and devices - know this: your world is changing, and it is changing faster than you think. In the past it was good enough [just] to know the hardware and software that you were working on inside and out. Those days are going away.
"So how can you help yourself, and what can you focus on to meet the requirements of a next generation engineer?”
2. 2. Learn Linux
“For years we focused on closed, monolithic operating systems, Cisco IOS is great example. But the times are changing: Nexus OS from Cisco is built in Linux, Arista is in Linux, and so is Cumulus Networks.
“Networking is moving in the direction of a Linux philosophy, so if you don’t know Linux start learning it now.”
3. 3. Learn to program
“The focus of the future is not on CLIs, it is on APIs. We need to consider how to manage systems as a whole, not as individual devices. So find a language and go for it. One of my goals to learn PowerShell, my team is actually knee-deep in Python.”
4. 4. Adopt ‘development’ philosophy
“It is not enough to learn the language - the philosophy behind programming and development are also important.
“You will hear the term infrastructure as code more and more often. The SDDC helps you down that path, but how can you help yourself? Learn about version control, repositories, configuration tools like Puppet and Chef, read up on lean and agile, use scrum. Take what you want from the programming world and make it yours.”
5. 5. Understand hypervisor technology
“The hypervisor is probably the most important technology development over the last 10 years. While it has disrupted the compute stack, know that network and storage are next – a combined $120bn market. The players that are the leaders in the markets today are not going to be the same in five years - disruption is happening before our very eyes and it going to be a wild ride.
“The hypervisor is not for the server guys any more, so go out and learn it, understand it - it will give you a huge advantage.”
6. 6. End siloed operations and collaborate
“Individuals or groups cannot be sole owners of technologies. We need to lop off the heads of the IT silos - the fiefdoms of the past must fall. Teams must collaborate: if you don’t make the change it will be forced upon you.
“Look for projects to collaborate with your peers on networking, computing, storage, operations, development and security. Create opportunities and network with them - invite them out for a beer. You are all in the same boat now.”
7. 7. Deep protocol knowledge
“Deep protocol knowledge is a fundamental for any good network engineer. You know how systems can communicate. In the past it was good enough to know one vendor’s implementation: tomorrow that will not work for you.
“You need to know how to turn the knobs on a system whether it is from Cisco, Palo Alto Networks, perhaps Facebook or even some startup team planning disruptive technology.
“Focus on the open protocols and learn them deeply. If you do so you put yourself in a position to add value to the business in ways you never thought possible. This creates conditions to innovate and move the needle.”